|
Accounting for over 80% of human-perceived information about the
physical world and as a foundational mechanism for autonomous
vehicles to "observe'' driving conditions, vision is a critical
window into the physical environment for both humans and engineered
systems. Natural human and machine vision, however, is subject to
inherent physical constraints such as being limited to the line of
sight (LOS). In road transportation, such vision constraints lead to
stress, inefficiency, and accidents in driving, and making turns
with obstructed views is a major cause for about 2.5 million
intersection accidents in U.S. every year. To transform human vision
and machine vision beyond the line-of-sight constraint, this project
proposes to leverage multiple visual sensors to enable humans and
engineered systems to see-through obstacles, thus transforming the
ways humans and engineered systems interact with environments. To
this end, this project develops the wireless networking and 3D
vision foundations for real-time wireless-networked augmented vision
which holds the potential to enable drivers and vehicles to see
through obstacles.
Eliminating the LOS constraint of natural human and machine vision
and enabling non-LOS surrounding sensing, the developed augmented
vision system will not only transform the experience, safety, and
comfort of human driving, it will also serve as an important
building block for human-in-the-loop autonomous driving and
fully-autonomous driving. The developed technologies are also
broadly applicable to domains such as public safety and disaster
response, thus having positive societal impact. This project also
integrates wireless-networked augmented vision research with the
cyber-physical-systems graduate programs at Iowa State University
and Wayne State University, and it uses augmented vision research to
enrich undergraduate education and research as well as K-12
outreach.
Plenary demo
& interview at 2017 US Ignite Application Summit
Publications (Selected):
- Y. Wang, G. Yan, H. Zhu, S.
Buch, Y. Wang, E. M. Haacke, J. Hua, and Z. Zhong, VC-Net:
Deep Volume-Composition Networks for Segmentation and
Visualization of Highly Sparse and Noisy Image Data, IEEE
Transactions on Visualization and Computer Graphics (VIS),
27(1), 2021.
- Ling Wang,
Hongwei Zhang, Analysis
of Joint Scheduling and Power Control for Predictable URLLC
in Industrial Wireless Networks, IEEE International
Conference on Industrial Internet (ICII), 2019
- Hajar
Hamidian, Zichun Zhong, Farshad Fotouhi, and Jing Hua, "Surface
Registration with Eigenvalues and Eigenvectors," IEEE
Transactions on Visualization and Computer Graphics,
2019.
- Sikai Zong,
Zichun Zhong, and Jing Hua, "Surface
Reconstruction by Parallel and Unified Particle-Based
Resampling from Point Clouds," Computer-Aided
Geometric Design (GMP
19), Vol. 71, pp. 43-62, 2019.
- Hai Jin,
Yuanfeng Lian, and Jing Hua, Learning
Facial Expressions with 3D Mesh Convolutional Neural Network,
ACM Transactions on Intelligent Systems and Technology, 10(1),
2019.
- Artem
Komarichev, Zichun Zhong, and Jing Hua, "A-CNN:
Annularly Convolutional Neural Networks on Point Clouds,"
IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
2019.
- Hongwei Zhang, Xiaohui Liu,
Chuan Li, Yu Chen, Xin Che, Feng Lin, Le Yi Wang, George Yin,
Scheduling with
Predictable Link Reliability for Wireless Networked Control,
IEEE Transactions on
Wireless Communications (TWC), 16(9):6135-6150, 2017
(A short version appeared in
IEEE/ACM IWQoS'15; ppt,
TinyOS
code] 16(9):6135-6150, 2017
- Shengbo Eben Li, Yang Zheng,
Keqiang Li, Feng Gao, Yujia Wu, Hongwei Zhang, J. Karl
Hedrick, Dynamical
Modeling and Distributed Control of Connected and Automated
Vehicles: Challenges and Opportunities, IEEE
Intelligent Transportation Systems, 9(3):46-58,
2017
Broader Impacts
(Selected):
- PI Zhang has shared project results and insight as a panelist
at the ``Innovation in the Smart Rural Ecosystem" Panel of 2019
NIST Smart and Secure Cities and Communities Challenge Expo, and
at the ``C-V2X for Future Automated Driving and Cooperative ITS"
Panel of the 2019 IEEE International Conference on
Communications (ICC).
- PI Zhang has developed a new graduate course "CPR E 548:
Cyber-Physical Systems Networking" at Iowa State University, to
address the unique networking needs of cyber-physical systems
and their applications in smart agriculture, smart
transportation, Industrial 4.0, and smart energy grid. The
course has been being offered regularly since fall 2018.
- PIs
Zhang Hua have led the establishment of the cyber-physical
systems (CPS) graduate program in the College of
Engineering, Wayne State University. Besides Wayne State
University, the program is highly supported by industry
partners such as Ford, GM, Lear, Magna, ODVA, and Automation
of Things.
- PI
Zhang has created and taught the new course ``CSC 5260/ECE
5260: Introduction to Cyber-Physical Systems" in winter 2017
that focuses on introducing the technology foundations of
cyber-physical systems to graduate students and senior
undergraduate students. This course examines a wide range of
topics including modeling, design, analysis, and
implementation of cyber-physical systems, dynamic behavior
modeling, state machine composition, concurrent computation,
sensors and actuators, embedded systems and networks,
feedback control systems, analysis and verification
techniques, temporal logic, and model checking. The course
helps prepare graduate and senior undergraduate students for
pursuing advanced topics in areas such as connected and
autonomous vehicles, Industry 4.0, Internet of Things (IoT),
and smart and connected health.
- PIs
Zhang and Hua have been actively training Ph.D. students
involved in the project, through regular research group
meetings as well as individual meetings and discussions.
|